Research map / Image generation / Editing and style transfer
Look here! A parametric learning based approach to redirect visual attention
Youssef A. Mejjati, Celso F. Gomez, Kwang In Kim, Eli Shechtman, Zoya Bylinskii
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| apply transforms to foreground as well as background (vs bg-only) (ablation) | GazeShiftNet · Mechrez | Absolute saliency increase: 0.54 -> 3.80 | Improved |
| apply transforms to background as well as foreground (vs fg-only) (ablation) | GazeShiftNet · Mechrez | Absolute saliency increase: 2.53 -> 3.80 | Improved |
| predict image-adaptive parameters instead of fixed mean parameters (ablation) | GazeShiftNet · Mechrez | Absolute saliency increase: 1.95 -> 3.80 | Improved |
| add tone and color adjustments to sharp+exp+cont (ablation) | GazeShiftNet · Mechrez | Absolute saliency increase: 1.47 -> 3.80 | Improved |
| add subtle transforms (sharp, exp, contrast) to tone+color (ablation) | GazeShiftNet · Mechrez | LPIPS Full: 10.97 -> 5.96 | Improved |
| add subtle transforms (sharp, exp, contrast) to tone+color (ablation) | GazeShiftNet · Mechrez | Absolute saliency increase: 5.33 -> 3.80 | Hurt |
| add saturation adjustment (ablation) | GazeShiftNet · Mechrez | LPIPS Full: 5.96 -> 9.70 | Hurt |
| add saturation adjustment (ablation) | GazeShiftNet · Mechrez | Absolute saliency increase: 3.80 -> 4.84 | Improved |
| use chosen operation order (sharp, exp, cont, tone, color) vs col,ton,con,exp,sha (ablation) | GazeShiftNet · Mechrez | LPIPS Full: 6.74 -> 5.96 | Improved |
| use GazeShiftNet instead of HAG | HAG · Mechrez | Absolute saliency increase: 2.24 -> 3.80 | Improved |
| use GazeShiftNet instead of GAT | GAT · Mechrez | LPIPS Full: 25.64 -> 5.96 | Improved |
| use GazeShiftNet instead of HAG for runtime | HAG · 30 high-res images | Average run time (s): 4743.54 -> 8.87 | Improved |
Closest papers by meaning
- Attention Distillation: A Unified Approach to Visual Characteristics Transfer (CVPR 2025 · similarity 0.86)
- FLATTEN: optical FLow-guided ATTENtion for consistent text-to-video editing (ICLR 2024 · similarity 0.85)
- Towards Understanding Cross and Self-Attention in Stable Diffusion for Text-Guided Image Editing (CVPR 2024 · similarity 0.84)
- VideoGrain: Modulating Space-Time Attention for Multi-Grained Video Editing (ICLR 2025 · similarity 0.84)
- Training-and-Prompt-Free General Painterly Harmonization via Zero-Shot Disentenglement on Style and Content References (AAAI 2025 · similarity 0.84)
- LookHere: Vision Transformers with Directed Attention Generalize and Extrapolate (NeurIPS 2024 · similarity 0.84)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
